Modularise editable element text access

[divec]

Move the code specific to each type of editable element text access into a
dedicated class, and use a factory pattern so a project using jQuery.IME can
register its own editable element text access implementation.

TextEntry,
TextEntry#getTextBeforeSelection( maxLength ),
TextEntry#replaceTextAtSelection( precedingCharCount, newText ):

• Abstract base class with methods to read/replace editable element text

FormWidgetEntry,
ContentEditableEntry:

• Concrete implementations for particular editable element types

TextEntryFactory,
TextEntryFactory#wrap( $element ),

• Factory class with method to wrap editable element

$.ime.TextEntry,
$.ime.textEntryFactory:

• Publicly accesible abstract base class and factory singleton

[santhoshtr]

Developers outside Wikimedia will not be familiar with OOJS. You can remove reference to OOJS and just state what really the method does- in this case having static members initizlied with empty object.

[edg2s]

I think it's worth mentioning that this is based on OOJS so in future we don't add code which may conflict with it, and therefore leave the door open to actually using it.

[santhoshtr]

Very good improvements. Thanks.

I get 5397 assertions of 6061 passed, 664 failed. while running tests in FF(Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Firefox/45.0 Iceweasel/45.0a2). There are some 8 tests that are failing in Chrome but they are real failures because of input method mappings. So rest of the failures seems to be a regression for Firefox.

A sample is given below. All CE tests return blank output.

Odia InScript ka -> କୋ - contenteditable
Expected:   

"କୋ"

Result:     

""

Diff:   

"କୋ" "" 

Source:     

imeTest/</<@http://localhost/jquery.ime/test/jquery.ime.test.js:308:1
x.Callbacks/c@http://ajax.googleapis.com/ajax/libs/jquery/1.10.1/jquery.min.js:4:26046
x.Callbacks/p.add@http://ajax.googleapis.com/ajax/libs/jquery/1.10.1/jquery.min.js:4:26361
imeTest/<@http://localhost/jquery.ime/test/jquery.ime.test.js:293:4

At the same time, when I tested manually in FF by typing something in the CE, I dont see any functionality broken.So I guess issue is in the testing instrumentation.

[divec]

Hmmm, I think there's something strange going on with the tests on Firefox. I'm seeing 53 failures on this commit (d6778c8), versus 693 failures(!) on the prior commit (ea97e4b).

Server: cd src/jquery.ime && python -m SimpleHTTPServer
Browser: firefox http://localhost:8000/test/index.html
User agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:42.0) Gecko/20100101 Firefox/42.0

[divec]

On a subsequent identical run on this commit ( d6778c8 ), I see only 8 failures on Firefox (the same 8 failures as on Chromium), so I think there's a race condition or something similar.

[santhoshtr]

Merged in commit 65cdfaf
Thanks!